Abstract We examined the processes of recolonisation and recovery of the benthic biota at nine subtidal stations 8 and 12 months following a bloom of the toxic dinoflagellate Karenia brevisulcata in Wellington Harbour, New Zealand. Compared with 2 months post-bloom, number of species (S) was significantly higher at three of six stations, numbers of individuals (N) increased significantly at four of six stations, species diversity (H¢) increased significantly at one of six stations and decreased significantly at one of six stations, and evenness (J¢) decreased significantly at three of six stations. At the three new stations for which we have data only for 8 and 12 months post-bloom, S increased significantly with time on one occasion, N increased significantly on two occasions, H¢ increased significantly on one occasion, and J¢ decreased significantly on one occasion. At all stations, changes in biomass (B) showed no significant trend, were all station and taxon-specific, and B was a poor indicator of recolonisation and succession. Biological communities became more similar between stations with increasing time after the bloom. Multi-dimensional scaling (MDS) analysis separated two stations with high levels of natural disturbance from the seven stations which were affected by the bloom. MDS analysis of pre- and post-bloom data for four stations indicated that the three stations most affected by the bloom followed trajectories of recolonisation and succession consistent with rapid but incomplete community recovery. Despite the widespread and devastating nature of the bloom, this event was only a minor perturbation to the ecology of soft substrate communities in the harbour.
Keywords benthic community ecology; subtidal soft substrate; mass mortality; recolonisation; succession; Wellington Harbour, New Zealand
New Zealand Journal of Marine and Freshwater Research, 2006, Vol. 40:
29–42
0028–8330/06/4001–0029 © The Royal Society
of New Zealand 2006
M04070; Online publication date 31 January 2006. Received 31 March
2004; accepted 11 August 2005
PDF file of entire paper: Print-quality (801K) | screen-quality (401K)